And to quickly piggy-back of Doug's point, I'm interest in more general lessons about crowdsourcing that Edmund and Shaf have learned from RED and are applying to READ-IT, since some of the other projects represented in the workshop are considering that approach.

@dougseefeldt - many thanks for acting as discussant, and starting off the discussions for panel 6. In answer to your questions:

In relation to interface translations, this is something we felt obliged to do, as READ-IT is a pan-European project, and the interface does need to be available (at a minimum) in all the consortium languages (Czech, Dutch, English, French, German, Italian). There is a basic argument of equality of access in operation here, as well of course, of facilitating contributions and responses from a range of users. Interface translation - just like agreeing ontology terms - raises all sorts of really interesting and complex issues around the ways in which we talk about books and reading in different languages/linguistic communities. One potential spin off from such an initiative is the development of further research topics or projects. A good example emerging from READ-IT might be a more systematic, comparative analysis of the lexicons of reading across a range of European languages over three centuries - you could systematically analyse (for example) the ways in which positive, negative or neutral terms are used in responses to reading; or examine metaphors associated with reading in different languages.

In terms of advice for fledgling projects, my suggestion is to build your network before bidding for a large scale grant, and have a good sense of how ways of working might be different in various institutions/countries. Always have plenty of project management or admin support as well (this is invaluable!)

Great question about the postcards: if harnessed and distributed as part of a campaign (e.g. tied in with an online or face to face event), the completion and return rate can be quite good, and certainly higher than for some forms of online surveys. One thing we have facilitated for UK based respondents is free return posting - so it costs them nothing to return the cards. Internationally, as the postcards also have the QR code on them, contributors can also return these digitally - just take a photo on your smart phone, scan the QR code, and you're ready to submit. As with all forms of public engagement, the success rate will depend on how accurately you identify your likely participant group, and how you can show that this benefits them. We've already worked on a small scale with a range of potential stakeholder groups, from high school students to residents in a secure unit, and it's worth trying this out and seeing what kind of engagement you can achieve. Local libraries, reading groups, cultural associations and bookshops will of course be obvious public engagement partners. Before the pandemic started, we already had quite developed plans for a university library in France to hand out a random postcard with each loan item, and have a physical drop box in their lobby for returns - which strikes me as a very easy, low cost (even no cost) engagement strategy.

The question about funding is a tricky one: I would suggest 'and/and' rather than 'either/or'. If your project can address or engage with a local issue, community or stakeholder group, by all means go for funding at the local level - but of course, for larger projects, you will certainly want national or international research council funding. It shouldn't at any rate, risk flattening out any analysis of results, because you can always structure data collection and curation to allow for both detailed micro-analytical work, and the broader macro-analytical questions. I think as historians of reading we tend to do both of these things anyway.

Posted by: @jim-connolly

And to quickly piggy-back of Doug's point, I'm interest in more general lessons about crowdsourcing that Edmund and Shaf have learned from RED and are applying to READ-IT, since some of the other projects represented in the workshop are considering that approach.

@jim-connolly - Great question! I think it's worth having a twin track strategy, because crowdsourcing can be either small in scale and highly intensive, or wider in scale but less time engaged. Both approaches will yield results.

For the UK-RED project, it was the brilliant @katiehalsey who really developed and delivered our initial volunteer recruitment and training system, with support from the rest of the team. We actively approached individuals and groups who we felt might contribute as volunteers: these were often retired librarians, lifelong learners, local history buffs, students, and members of author societies. We ran roadshows to demonstrate the database and recruit volunteers; Katie helped put together and send out a regular newsletter to volunteers; and we took flyers to conferences and events, and harnessed listservs, to recruit new volunteers (this was in the age before social media!) Volunteers were given an individual training session and mentoring, and because of the way in which data was collected and checked by project team members, one of us was always on hand to check the accuracy of volunteers' work. We always tried to match up volunteers with their area of interest - the key approach for UK-RED was to make volunteers feel that their contribution was useful and meaningful for them, as well as the wider public (and the project). For UK-RED, during the data collection period, in total over 120 volunteers contributed some 6,000 entries, so while it's not a huge crowdsourcing effort in terms of numbers, it is very impressive in terms of depth of engagement. Several volunteers went on to write PhD theses or monographs on the basis of their contributions. I would describe this as targeted and very deep crowdsourcing; a select community of volunteer contributors was identified, supported, and made to feel part of a collaborative research effort.

For READ-IT, we have also been casting our net wider, and developing strategies for broader crowd contribution, based on a more digital social media savvy population. The chatbots for example, are designed to facilitate very quick, short, but accurate responses from the public, without them having to do very much: just reply to a few prompt questions as and when they feel like it. The conversion rate from access to contribution will necessarily be smaller in this case - it's usually around 5% - but the infrastructure is cheap to run and deploy, and you can make it work at scale without any issues. The fact that data gathering is anonymous and GDPR compliant is an added bonus.

In summary, be sure to develop crowdsourcing strategies that are both 'light touch' and those which are 'deep' and highly invested in the project; identify your potential contributor communities, and make sure there is something in it for them. 

Posted by: @dougseefeldt

nts:

My first question is directed at Edmund. UK RED is a terrific example of a digital humanities project with a long horizon as it dates back to the mid-1990s when the Web was in its infancy (much like printing at the beginning of the project's chronological period). What can new or even nascent projects learn from UK RED's idiosyncratic development history? One would not recommend a new project follow this exact same path, would one?

Many thanks for this question, @dougseefeldt! I indicated a few things in my presentation that stood out, one of the most important of which (it seemed to me) was people. Really, it was the postdocs and postgrads who were associated with the project that were responsible for much of the energy and direction of research. I'm thinking here of Stephen Colclough's early work locating and defining archival source material for his research period (the early nineteenth century), or Ros Crone bringing her crime historian's eye to bear on reading history and opening up a whole vista of potential evidence in both court reports and prison records. Developers who have an interest in both programming and the humanities and can find innovative ways of bringing the two together. And of course the leadership of Directors has been hugely important: seeking out and working with collaborating partners in other institutions; drafting and writing bids; securing funding. All of this seems fairly obvious, but in DH there may be a tendency to focus on the technology when the human, team, and institutional elements are extremely important.

Of course, any "new or ... nascent projects" necessarily won't be developed in the same university, funding, or indeed internet environment that RED emerged from. The particular "conditions of digital possibility" I outline in the presentation are inescapably different from those that now hold sway. So, what I'm gesturing at here, I guess, is: how do we apply a history of ideas approach to DH and DH legacy projects? I guess these would be the same sorts of questions that Willard McCarty has been raising for a number of years about humanities computing.    

Thanks for the kind acknowledgement, @stowheed! Working with the volunteers while a postdoc on RED remains one of the most fascinating and rewarding experiences of my career - and I learned a huge amount about keeping a volunteer community motivated, engaged and interested. I loved corresponding with, and in some cases, meeting, our volunteers, who came from a range of really interesting backgrounds, and who were without exception interesting and dedicated people. In fact, that work has borne fruit in unexpected ways - one of the 'serendipitous evolutions' that Kyle Roberts was discussing in his talk in Panel 2 is that a former RED volunteer, with whom I'd worked closely on entering marginalia from two collections in North Eastern Scotland, got in touch when we launched the 'Books and Borrowing' project to ask whether we'd like to include one of those libraries as a project partner. Not only did she broker the arrangements with the owner of the private library, she even donated her own transcription of the borrowers' register there! I find myself utterly humbled by the kindness of members of the volunteering community, who give so much to projects like ours. 

But the kind of deep and targeted crowdsourcing that we were doing at RED is difficult to manage - Shaf mentions the importance of editorial oversight in the contributions we received from RED volunteers, and that was of course enormously important.  But that's only one element of it. The major element, in my experience, was time, and I'd strongly advise anyone planning on including this kind of thing in their project to set aside a substantial portion of time dedicated to managing this, and then treble that amount of time. Of course all successful relationships depend on putting the time and effort in to the relationship, but I think this is particularly so when substantial quantities of goodwill are at stake (as they always are in a volunteering situation).  

I'm interested to see the different approach taken by READ-IT, and very much look forward to finding out more about how this works!